Impulse Lines For Differential-Pressure Flow Meters.
Impulse Lines For Differential-Pressure Flow Meters.
Impulse Lines For Differential-Pressure Flow Meters.
www.tuvnel.com
Impulse Lines for Differential-Pressure Flow Meters
Executive Summary
This guide presents good practice in the design and installation of impulse lines (small bore pipes) that connect a
differential-pressure flowmeter to the instrument for measurement of the pressure difference. It is intended to assist
the designer to avoid known problems with impulse lines that can lead to incorrect measurement. Examples of typical
problems with impulse lines are gas bubbles trapped in a line intended to be filled with liquid, condensation in a line
intended to be filled with gas, pulsations in flow, temperature effects, and blockages.
The design of a flow measurement installation using differential-pressure flow meters involves consideration of a number
of issues related to the impulse lines. Geometrical aspects include the impulse line diameter, length and slope and the
detailed routing. The location of the pressure tappings around the pipe circumference must be defined in conjunction with
the selection of the impulse line fluid. Other issues concern isolation valves and other fittings, and the implications of the
working fluid and ambient temperatures.
The document presents detailed guidelines on these aspects, as well as specific recommendations for wet gas and
steam applications.
A further section considers design for pulsating flow conditions, where the impulse lines can distort the mean
differential pressure.
In general, the recommendations are to use impulse lines of equal length and as short as practicable, to use recommended
configurations, to use trace heating and/or insulation if required, to ensure that the temperature is the same in both
impulse lines, and to ensure effective venting of trapped gas in a liquid-filled line and draining of any liquid in a gas-filled
line. Moreover, if the flow is pulsating it is also recommended that changes in impulse-line diameter should be avoided and
that the pulsation and the impulse-line frequencies should not coincide.
Contents
Executive Summary 1 3. Pulsating Flow 11
1. Introduction 2 4. Impulse-line Blockages 14
2. General Principles and Guidelines 5. Two-phase Flow 15
2.1 Introduction 3 6. Impulse-line Purging 15
2.2 Geometry 3 7. Close Coupling 16
2.3 Fluid 5 8. Summary 17
2.4 Temperature effects 5 9. References 17
2.5 Location of secondary instrument 5
2.6 Location of pressure tappings 6
2.7 Impulse-line installations 6
2.8 Specific guidelines for 9
Wet-gas applications
2.9 Specific guidelines for steam 9
2.10 Condensing or evaporating fluids 11
1
Good Practice Guide
1. Introduction
An impulse line is a small-bore pipe that is used to connect a point in a pipe at which pressure is to be measured to an
instrument. In flow measurement using a primary device such as an orifice plate, a nozzle, or a Venturi meter, impulse lines
are used to connect points upstream and downstream (or in the throat) of the meter to a secondary device for measuring
the differential pressure.
The use of impulse lines is known to cause a number of problems that can lead to an incorrect measurement. The
problems can arise from the following:
1. Damping of the pressure signal or resonances (this problem is exacerbated if a transient measurement is required
or if the lines are of different length)
2. Blockage
3. Leakage at couplings
4. Different temperatures (and therefore different densities) in a pair of impulse lines
5. The fluid in the impulse lines being of a different composition from the fluid in the pipe (this causes a particular
problem in a vertical meter when a gravity head correction or some other approach is required)
6. Condensation in an impulse line that is intended to be filled with gas
7. Gas bubbles being trapped in an impulse line that is intended to be filled with liquid, or boiling of a liquid with a
bubble point below the ambient temperature.
Gas trapped in
liquid system
Damping in
Exposure to
long lines
different
temperatures
Leaks at joints
Liquid trapped
in gas system
2
Impulse Lines for Differential-Pressure Flow Meters
1. Introduction cont.
This document provides general guidance for the design and installation of impulse lines. It covers impulse lines but not
the pressure tappings (their diameter, edge sharpness etc.) themselves. It applies to a range of applications including liquid,
gas, two-phase, steady and pulsating flows. It does not cover safety issues. However, in general appropriate materials,
manufacturing techniques, and sealing should be used, and applicable standards should be followed.
The key international standards relating to flow measurement using differential-pressure meters and to impulse lines are
ISO 51671 and ISO 21862 respectively. Further information on flow measurement can be found at www.tuvnel.com.
2.1 Introduction
The design of a typical differential-pressure flow-measurement installation involves consideration of a number of issues
related to the impulse lines:
These issues are considered below for steady flows. Special considerations that apply to pulsating flows are discussed in
Section 3.
2.2 Geometry
Most of the problems described above can be minimized or avoided by the use of impulse lines that are as short as
possible. The main limitations are the possible need to isolate a secondary device from a high-temperature fluid in
the pipeline and the need to include valves between the primary and secondary devices to make possible removal or
replacement of the secondary device and to make it possible to open a path between high- and low-pressure sides with
no differential pressure. A typical arrangement is shown in Figure 2. Additional valves, often in a single manifold, may be
required to permit draining, venting or calibration.
3
Good Practice Guide
Figure 2: Typical valve arrangement to isolate the secondary device and to provide equalization
Similarly, it is usually best to select the smallest possible impulse-line internal diameter, but this will be limited by practical
constraints such as avoiding damage and blockages, capillary effects, and the need to avoid trapped bubbles in liquid lines
and trapped liquid in gas-filled lines. Generally impulse-line diameters in the range 4 mm to 25 mm are used. The smallest
sizes are used for unsteady flow and for research. The international standard ISO 21862 recommends an internal diameter
of not less than 6 mm and preferably at least 10 mm if condensation is likely to occur or if gas bubbles may be liberated
from a liquid. It recommends a minimum internal diameter of 10 mm for industrial process applications in which reliability
is the primary concern. Generally it recommends a maximum internal diameter of 25 mm.
ISO 2186:20072 provides a table giving guidance on impulse line diameters for long impulse lines. The table, shown as
Table 1 below, is in an informative (i.e. not a normative) annex. It is introduced with the words: It is always recommended
that the shortest possible impulse line lengths be used. Where it is not possible to conform with this, guidance on the
preferred line diameter may be obtained from Table A.1 in ISO2186:2007 (which is equivalent to Table 1).
If long lengths of impulse piping are required, it is important to ensure that the piping is well supported so that the
possibility of damage due to vibration is eliminated. The natural frequency of the longest unsupported length should not
coincide with any potential exciting frequencies in the system.
4
Impulse Lines for Differential-Pressure Flow Meters
2.3 Fluid
In many applications, such as flow measurement of a gas that does not condense at ambient temperatures, the fluid in
the impulse lines is the metered fluid. In other cases, such as measurement of steam flow where the steam saturation
temperature is well above the ambient temperature, it may be advantageous for the fluid in the impulse lines to be in the
liquid phase.
It is essential to avoid gas or vapour bubbles in liquid-filled lines, and liquid in gas-filled lines.
In some applications, for example steam flow with condensate (here meaning liquid water) in the impulse lines, it may be
necessary to fill the impulse lines prior to start up.
The pressure head in an impulse line depends on the fluid density, which is in turn affected by the fluid temperature. It
is therefore good practice to ensure that both impulse lines are exposed to the same ambient temperatures. For example
it would be bad practice to have one impulse line exposed to solar radiation and the other in shade. For this reason it is
recommended that the impulse lines are tied together, even when they are insulated.
Insulation and protection from solar heating is particularly important if the impulse lines contain a liquid that is close to its
bubble point. It is also desirable to use insulation to minimize fluctuations in ambient temperature, which could affect the
stability of the pressure head in the impulse line.
For example, suppose the impulse lines to a differential-pressure transducer contain water and are at temperatures of 20 C
and 30 C. If there is a difference in height between the pressure tappings and the transducer of 2 metres, there will be an
error in measured differential pressure of 49 Pa.
For gas-filled impulse lines the secondary instrument should be located above the primary instrument to facilitate drainage
of any liquid that may form in the lines. Similarly the secondary instrument for liquid-filled lines should be located below
the primary instrument to facilitate venting of gases.
With liquid-filled lines it is essential that the liquid heads in the two impulse lines are equalized or otherwise compensated.
One way of achieving this where steam flow is being measured is described in Section 2.9.
If oil is being measured and there is the possibility of contamination with water, then the classic arrangement of the
transducer below the tappings may not work well since water may enter the impulse lines. In this case it may be better
to place the transducer above the tappings and to make additional provision for the possibility of gas in the impulse lines.
This problem is not specific to oil and water but may occur with other immiscible pairs of liquids.
5
Good Practice Guide
The recommendations of ISO 2186:20072 for horizontal pipes are summarized in Table 2. The underlying rationale is that
gas- or vapour-filled impulse lines should slope upwards to facilitate drainage of liquid, and that with liquid-filled lines
tappings below the centre line can accumulate solids while tappings above the centre line can accumulate non-condensing
gases.
Table 2: Recommended pressure tapping locations from ISO/CD 2186:2004 for horizontal meters
With steam in horizontal pipelines, tappings between the horizontal centre line and 45 below the horizontal may also be
acceptable.
In vertical pipelines there is no restriction on the radial location of the pressure tappings.
The recommendations above on the location of the secondary instrument relative to the primary instrument imply that gas-
filled impulse lines should slope upwards from the pressure tappings to the transducer and liquid-filled lines should slope
downwards. ISO 2186:20072 states that impulse lines shall be installed such that the slope is in one direction only, and that
if a change in direction of slope is unavoidable, only one such change is made. In this event a liquid trap is required at the
low point and a vapour vent at the highest point. The minimum recommended slope for self-draining or self-venting is 8
per cent.
Examples of impulse lines for dry natural gas are given in Figures 3 and 4. Figure 3 shows impulse lines leading from an
orifice plate with horizontal tappings to an instrument room. The lines are sloped upwards towards the instrument room,
and are insulated.
6
Impulse Lines for Differential-Pressure Flow Meters
Figure 4 is a view of the interior of the instrument room. It shows the impulse lines again correctly sloping upwards
towards the differential-pressure transducers so that any liquids that form in the lines will tend to drain downwards. Since
the cabin is at a constant temperature which is above the dew point of the gas in the impulse lines, there is no need for
insulation. The transmitters are calibrated in situ.
Many instrument enclosures are much smaller than the room in Figure 4. Figure 5 shows an enclosure with three
differential-pressure transmitters and one static-pressure transmitter. The enclosure contains a heater (behind the
transmitters in the middle of the picture). The three differential-pressure transmitters are mounted on a mono-block.
Where high-static calibration is performed off site and footprinting is the norm on site then small enclosures are not a
problem since if the transmitters are mono-block-mounted the complete block with transmitters mounted is removed for
remote calibration, and footprinting can be performed in situ on site with very little problem. The mono-block also has a
stabilizing effect with regard to temperature. Where a number of transmitters are connected via pipework it is prudent
to keep the distance between transmitters as small as possible. On some gas systems where three or more transmitters
are used and the distance between the transmitters is about 1 m there is a problem with gas noise; the signal from the
end transmitter may be reflected so as either to increase or to decrease the signal strength at the first transmitter. This
causes problems where the accuracy of the measured differential pressures is checked by a comparison between pairs of
transmitters, and alarms are set. Thus close coupling between such transmitters is very desirable (see Figure 5).
7
Good Practice Guide
The recommended slopes of the impulse lines for meters in vertical pipes are generally the same as for horizontal meters.
If liquid-filled lines are used to measure flow of a vapour or gas in a vertical pipe, there is a difference in liquid head
between the tappings. Thus there is a pressure difference between the tappings under static conditions for which
compensation can be made in one of two ways:
1. Form the lower impulse line upwards before turning horizontal at the same height as the upper impulse line, and
then form both lines downwards to the secondary device. This arrangement equalizes the liquid head above the
secondary device in the two impulse lines and avoids the need for a correction to the instrument calibration.
2. Both impulse lines leave the tappings horizontally and then turn down to the secondary device. The zero of the
secondary device must then be adjusted to account for the difference in liquid head above the secondary device.
To Dp
measurement
Figure 6: Handling the height difference between tapping points on vertical flow
ISO 2186:20072 provides recommendations on impulse-line configurations with useful diagrams and should be the first
place where information is sought. The standard has a useful section on isolation valves, which are required in practice,
though their use conflicts with the desirability of having as few joints as possible in the impulse lines and having no
changes of diameter. Some choices of valve are better than others: globe valves with a vertical stem that may create a
pocket of gas or liquid should be avoided. It is usually desirable to be able to join the high- and low-pressure sides of the
differential-pressure transducer so that the signal from the transducer can be obtained with no differential pressure but at
line pressure (see Figure 2); this is often called performing a high-static zero check.
Installations other than those recommended can be used, but will generally require more maintenance. If the transducer
lies below the pressure tappings in gas, drain pots and valves will be required; if the transducer lies above the pressure
tappings in liquids, vent valves at least will be required.
8
Impulse Lines for Differential-Pressure Flow Meters
Some installations may require the opportunity to rod out the impulse lines. This is particularly important in steam flow
as dirt is often a significant problem. In addition to the requirements in the previous paragraph, provision for draining
or venting the impulse lines may be necessary, both immediately after installation (e.g. after hydrostatic testing) and
subsequently. The specification of the piping for the impulse lines may be different from that for the process piping
because of the more limited temperatures in the impulse lines.
For steam flow, plugged tees in the impulse lines may be required to enable the impulse lines to be filled with water at
start up.
The UK Department of Trade and Industry (now DECC) Licensing and Consents Unit Guidance Notes for Petroleum
Measurement under the Petroleum (Production) Regulations December 2003, Issue 7 provide specific comments for wet
gas applications.
A particular concern is that liquid drop-out in impulse lines is likely to occur due to cooling of the gas towards ambient
temperatures once it leaves the meter stream. In extreme cases hydrates may form. To avoid measurement errors due to
the presence of either liquids or hydrates in impulse lines, the Guidance Notes reiterate the general recommendations that
impulse lines should be as short as possible and inclined upwards in order to drain entrained liquids. Other advice given
in the Notes is that liquid or hydrate accumulation can be further countered by the insulation of the impulse lines and the
application of trace heating, and that operators should consider placing the transmitters, and even the impulse lines if
possible, in a heated, sealed enclosure.
The Notes advise that catchment pots located in the impulse lines may be effective at catching liquids, and that these
should be drained frequently to avoid excessive liquid build-up. They may therefore not be effective on un-manned
installations that experience significant liquid drop-out in impulse lines.
It is also recommended for wet gas applications that the meter should be orientated horizontally such that the pressure
tappings are located at the 12 oclock position. In this way the potential for liquid becoming entrained in the tapping or
impulse lines is minimized, and the tappings are kept as far as possible from the bulk of the liquid if stratified or annular
flow regimes are present.
Specific recommendations for steam flow are provided in A Review of Steam Flowmetering Technology3. It is reported
that impulse lines are usually designed such that they are filled with condensate at all times. This is usually achieved by
a siphon arrangement or by including a low point in the impulse lines in which the condensate collects, insulating the
transducer from the steam. Seal pots (also known as condensate pots) are often included in the arrangement and there is
usually provision for venting or flushing the lines and filling with water.
It is recommended in the report that impulse lines should ideally be insulated and provided with trace heating to avoid
condensate freezing in cold weather. Sometimes glycol is added to the condensate in the impulse lines for similar reasons.
9
Good Practice Guide
Gas bubbles in the impulse lines can cause a sluggish response from the pressure transducers and the impulse lines should
be purged to ensure that no air is trapped in them.
Some manufacturers also recommend the use of condensate or seal chambers for steam flow. They are used when the
instrument displaces a large volume of liquid as the measurement changes, which was much more common with older
instruments. A seal chamber is a small pressure vessel that is mounted at the top of the liquid-filled impulse line. Each
chamber acts as a reservoir in the impulse line in which large volume changes will result in minimal elevation change so that
seal liquid is not dumped into the process line and elevation shifts of the wet-leg liquid do not cause measurement errors.
The two seal pots must be at the same height. In general modern transducers with very small volumetric displacement
do not need seal chambers. Spirax Sarco do not recommend the use of condensate pots except for superheated steam
applications. Also, they do not recommend the configuration shown in Figure 6 and instead they recommend removing
the static-head offset using the secondary instrument.
To secondary
element
Figure 7: Illustration of the use of seal pots in steam flow
Figure 8, taken from Reference 3 (see also Mottram et al.4), shows some common arrangements for steam flow in
horizontal and vertical pipelines. Figure D.5 of ISO 2186:2007 is also very helpful.
10
Impulse Lines for Differential-Pressure Flow Meters
In flow measurement of a fluid that is close to its bubble or dew point, the general aim is to ensure that both impulse lines
are maintained full of either gas or liquid at all operating and ambient conditions.
The recommendation in Section 2.8 for steam flow applies to any condensable vapour for which the dew point is above
the ambient temperature, in which case impulse lines will tend to fill naturally with liquid. In other applications alternative
options may be appropriate depending on the dew and bubble points and the ambient temperature, on the mass fraction
of the condensable vapour and on whether or not trace heating or cooling is employed. For example, if there were the
possibility of condensation of only a small amount of a vapour component it would be better to follow the gas-flow
recommendations.
If the ambient temperature is always above the dew point, the recommendations for gas flow should be followed.
Flow measurement of a liquid that evaporates at ambient temperatures can present similar problems. It may be necessary
to provide insulation to prevent boiling of liquid in the impulse lines. Alternatively, if ambient temperatures are well above
the bubble point it may be better to design for vapour in the impulse lines.
3. Pulsating Flow
where is the time-mean value of the differential pressure
is the fluctuating component of the differential pressure
and is the root mean square value of
With a pulsating flow the requirement may be to measure either the time-dependent flow or a time-average.
The occurrence of flow pulsations introduces the possibility of additional measurement errors. An extensive research
programme on impulse lines in pulsating flows was performed at Southwest Research Institute in the USA, and another
major research contributor was the research laboratory of the Nova Chemical Corporation.
In the SwRI Research Programme, the report by Sparks5 describes results of an experimental programme to document
and to describe pressure-measurement errors inherent in impulse lines when gas flow in the pipe is unsteady or pulsating.
The report contains good descriptions of the effects of impulse lines on pressure measurements in pulsating flows. It is
directed towards measurement of the time-average flow.
11
Good Practice Guide
With a simple impulse line consisting of a straight tube of constant diameter its resonant frequencies will be excited when a
pulsation frequency coincides with its quarter-wave resonant frequency or with any odd multiple of that frequency. For this
simple geometry the acoustic resonance frequencies fn are given by
where
The experiments showed that at resonance the pulsation amplitudes in the impulse lines can be much greater than in the
main pipe.
Further analysis showed that pulsation data from impulse lines cannot necessarily be relied upon unless impulse-line lengths
are very much shorter than a quarter wavelength for the highest frequency existing in the piping, and then only if there are
no constrictions or volumes in the impulse line that would lower its resonant frequency.
Moreover, the existence of pulsations can cause a shift in the mean pressure inside an impulse line. The average pressure
transmitted by the impulse line is not necessarily equal to the average pressure in the piping system, even when impulse
lines are not at resonance. The basic cause of the shift is stated to be oscillating flow through non-linear resistance
elements such as contractions, enlargements, valves and transducer volumes. These elements may have a loss coefficient
for flow into the impulse line that is different from the coefficient for outward flow. Oscillating flow is therefore partially
rectified, and a net change in the mean pressure in the impulse line results. This flow rectification is greatly amplified if the
impulse-line acoustic resonance frequency coincides with the pulsation frequency.
The acoustic response of the impulse lines can cause the pulsations in the impulse line to be many times higher or
lower than in the main pipe, depending on the acoustic response of the impulse line.
Impulse lines can distort mean-pressure measurements when there are pulsations in the piping system.
These shifts can be either positive or negative, and are most severe when acoustic resonances are excited in the
impulse line.
Pinched valves, volumes and impulse-line filters can be used to provide damping and therefore attenuate the
observed pulsations, but they do not totally eliminate the mean pressure shift.
Differential-pressure measurements are particularly susceptible to impulse-line effects. A positive mean-pressure
shift in one line and a negative shift in the other can cause a large shift in the differential pressure.
The main recommendation of the report was that impulse lines should be made as short as possible to give a high acoustic
frequency, well above any pulsation frequency in the main piping. The lines should also be of constant diameter, and be
free of constrictions and volumes.
12
Impulse Lines for Differential-Pressure Flow Meters
In the Nova Corporation Research, Botros et al6 made similar recommendations on impulse-line design to those arising from
the SwRI research, i.e.
ISO/TR 33137 provides guidelines on the effects of flow pulsations on flow meters. It encapsulates and adds to the
impulse-line design guidelines in the literature described above, but distinguishes between the use of a slow-response and
that of a fast-response transducer. A slow-response transducer is designed to indicate the time-mean differential pressure,
whilst a fast-response transducer will be designed to follow the time-dependent differential pressure so that the square root
of the measurement can be averaged to yield a true mean flowrate (i.e. the square-root error is eliminated).
Design guidelines for impulse lines for slow-response transducers to measure time-mean flowrate are:
A piezometer ring should not be used. A piezometer ring is also called a Triple-T arrangement and is shown in
Figure 1 of ISO 5167-1:20031.
The distance between pressure tappings should be small compared with the pulsation wavelength.
The impulse lines should be as short as possible, and of the same bore as the tappings, which should be greater
than 3 mm in diameter.
An impulse-line length close to the pulsation quarter wavelength should not be used.
For gas-filled lines the sensor cavities or other volumes should be as small as possible.
Vent points are required for liquid-filled lines to remove any gas bubbles.
Damping resistances in the connecting tubes and sensing element should be linear. Throttle cocks should not
be used.
The device time constant should be about ten times the period of the pulsation cycle.
If the above rules cannot be observed the secondary measurement system can be effectively isolated from
pulsation by the insertion of identical linear-resistance damping plugs into both impulse lines, as close as possible
to the pressure tappings.
Design guidelines for a fast-response transducer to measure the time-varying flowrate are:
The mechanical and electronic frequency limits of the secondary measurement system should be at least ten times
greater than the pulsation frequency.
The impulse-line lengths should be as short as possible and less than 10% of the pulsation quarter wavelength.
The bore of liquid-filled impulse lines should be greater than or equal to 5 mm.
Fittings and valves should be of the same bore as the impulse lines.
The secondary device must be geometrically identical on the upstream and downstream sides.
Vent points are necessary for liquid-filled lines to remove any gas bubbles.
13
Good Practice Guide
4. Impulse-line Blockages
Blockage of an impulse line will cause flowrate information to be lost, with potentially serious consequences in a process
control application. The need to avoid blockages is a primary consideration in the specification of impulse-line diameter.
Blockage of impulse lines due to freezing is addressed by Daiber and Hughes8, who estimate that 60% of trace-heating
systems for freeze protection are not working as designed. They found a number of problems with installation, and
transposition of the problems into the corresponding solutions yields:
1 Provide trace heating (generally by electricity or steam) whenever freezing conditions are possible.
2 Avoid exposure of trace-heating systems to excessive temperatures.
3 Avoid over-tightening or the improper clamping of the tubing bundle (consisting of the impulse lines and heating
line) to its supports, which can cause the insulation to be compressed and prevent proper operation of the trace-
heating system.
4 Do not exceed the tubing bend-radius limit.
5 Install effective trace-heating systems and insulation. The simple method of laying on insulating tape, applying an
electric heater cable and then adding more insulation tape and sealing with mastic may not provide adequate
protection.
6 Ensure correct slope of impulse lines.
7 Provide high-quality instrument enclosures, which, though expensive, are essential for reliable long-term frost
protection.
8 Provide adequate weatherproofing.
The use of smart pressure transducers to detect and predict plugged or frozen impulse lines is described by Menezes9.
An open impulse line has high frequency noise due to fluid turbulence. If a line plugs its signal becomes flat. A smart
transducer can learn the frequency characteristics of open impulse lines and use them to establish an OK condition.
The different frequency characteristics of plugged lines can then be detected and reported using software. It is possible to
detect the plugging of one or both lines. This technique requires the use of a very fast pressure sensor.
Tests to assess the ability of smart pressure transducers to detect an impulse-line blockage in a refinery application are
described by Szanyi et al10. Blockage in the impulse lines of pressure or differential-pressure transducers in fluid catalytic
cracking (FCC) units can cause very expensive shutdowns.
Advice on trace heating using steam is provided by Harrold11. When trace heating is used, both lines must be equally
protected. Harrold recommends avoiding splitting a single steam trace line into two lines, one for each impulse line, and
then rejoining the lines ahead of the steam trap. Steam follows the path of least resistance. The line offering the most
resistance will stop flowing, the steam will condense, freeze, and possibly rupture the trace-heating line. Its okay to split
the lines, but provide each line its own steam trap.
Trace heating must not be so great as to cause liquid to vaporize unintentionally or to prevent steam from condensing.
Waxing is another cause of blockage. Lagging or trace heating may be required to prevent this problem.
14
Impulse Lines for Differential-Pressure Flow Meters
5. Two-phase Flow
With two-phase flows the impulse lines are generally designed to be filled with either liquid or vapour depending on the
nature of the flow. Thus all of the comments in Sections 2 and 3 above apply equally to two-phase flows.
The location of the pressure tappings should take into account the probable two-phase flow pattern. For example with a
stratified flow and gas-filled lines, the tappings should be located on the top of the pipe.
In addition it is important to avoid condensation in vapour-filled lines and boiling in liquid-filled lines; so trace heating or
insulation may be required.
A pulsating two-phase flow will cause particular difficulty because of, for example, periodic pumping of gas bubbles into
liquid-filled impulse lines.
A meter located in a vertical pipe length requires particular care because the fluid in the meter line between the two
tapping points has a different density from the fluid in the impulse lines. The correction for the difference in height
between the tapping points requires a calculation of the mean two-phase density in the pipe line, which is subject to some
uncertainty. Moreover, although it is intended to have a known single-phase fluid in the impulse lines, in practice often the
fluid in the impulse lines is in fact two-phase. For this reason it was found in Reference 12 that a horizontal orientation of
a Venturi tube was preferred in multiphase flow.
The difficulties in maintaining a single-phase fluid in the impulse lines in some two-phase flow applications where, for
example, the dew or bubble point is close to ambient temperatures, may mean that it is preferable to have a horizontal
meter with the secondary instrument level with the primary instrument. This arrangement minimizes the uncertainties
associated with the static head between the tapping points and the secondary instrument.
6. Impulse-line Purging
One solution to problems of impulse line-blockage, gas bubbles in liquid lines and liquids in gas lines is to maintain a
continuous purge flow through the lines. Impulse lines can be purged with a dry gas to keep liquids or vapours out, and
with a non-freezing liquid to keep the process fluid out. It is important to purge through equal lengths of the two lines and
to equalise the two purging flowrates. Figure 9 below shows a typical installation with continuous impulse-line purging.
Figure 9: Impulse line installation with purge flow (F represents a flow meter to measure the purge flowrate)
15
Good Practice Guide
However (see Harrold9) there are disadvantages in terms of the requirements for additional hardware and piping and the
need for tight control of the purge flowrate. It will be necessary to ensure that the frictional pressure drop due to the
purge flow is negligible, and this may mean selecting a larger internal diameter of the impulse line.
Continuous liquid purging has been recommended as a method of avoiding undesirable effects of pumping of vapour
bubbles into liquid lines in pressure measurement in two-phase flows, especially in research applications.
7. Close Coupling
Recently some manufacturers have developed transducer mounting arrangements so that the differential-pressure
transducer is mounted directly on to the meter pipe. This removes the need for impulse lines in the conventional sense,
although there is still a length of tapping at least between the pipe and the differential-pressure transducer. This
arrangement has significant benefits in that it greatly simplifies the piping and, most importantly, reduces the number
of joints, and thus the number of leak paths at threaded connections and compression fittings. It reduces the need for
insulation or trace heating and errors due to different densities of fluid in the different lines; it also minimizes problems
with liquid drop-out or gas break-out in the impulse lines where the ambient temperature is very different from the line
temperature. If short impulse lines are required because of pulsations, close coupling gives the shortest available lines.
Close coupling also gives a good speed of response. However, it means that the transmitters have to be located in small
enclosures beside the meter rather than in large cabins further away. This may make it more difficult to calibrate the
transmitter in situ and may be less attractive to the operator. The small enclosures may provide shelter but not temperature
control; however, modern transmitters are less sensitive to changes of temperature than earlier models. The use of
impulse lines may be essential when the line temperature is too high or too low for the pressure transducer, or when
ambient conditions mean that the transmitters are better housed in an instrument room with a controlled environment
16
Impulse Lines for Differential-Pressure Flow Meters
8. Summary
In general the recommendations are to use short impulse lines of equal length (close coupling is a way to achieve this), to
use recommended configurations, to use trace heating and/or insulation if required, to ensure that the temperature is the
same in both impulse lines and to ensure effective venting of trapped gas in a liquid-filled line and draining of any liquid in
a gas-filled line. Moreover, if the flow is pulsating it is also desirable to avoid sudden changes in diameter and it is essential
to avoid coincidence of the pulsation and the impulse-line frequencies.
9. References
1 BRITISH STANDARDS INSTITUTION. Measurement of fluid flow by means of pressure differential devices inserted in
circular cross-section conduits running full. Part 1: General principles and requirements. BS EN ISO 5167-1:2003.
2 INTERNATIONAL ORGANIZATION FOR STANDARDIZATION. Fluid flow in closed conduits Connections for pressure
signal transmissions between primary and secondary elements. ISO 2186:2007.
3 A Review of Steam Flowmetering Technology. Report 2004/69 for TUV NEL Ltd, East Kilbride, Glasgow.
4 MOTTRAM, R.C., OWEN, I. and TURNER, J.T. Wet steam/wet gas flow measurement. Flow Measurement and
Instrumentation Consortium Report No 13, May 1991.
5 SPARKS, C. R. The effects of gage lines on pressure measurement accuracy. SwRI Report No. 84-6A. San Antonio,
Texas: Southwest Research Institute, May 1984.
6 BOTROS, K. K., JUNGOWSKI, W. M., and PETELA, G. Gauge line effects and DP transmitter response to flow
pulsation through orifice plate. Flow Measurement and Instrumentation, 3 (3), pp. 130-144, 1992.
17
Good Practice Guide
9. References cont.
8 DAIBER, J. and HUGHES, A. Freeze Protecting Your Power Generation Assets OBrien Corporation, (http://www.
babcockpower.com/pdf/vpi-14.pdf).
9 MENEZES, M. Improving plant safety & availability through advanced measurement diagnostics. Originally
presented to ISA/2000 (August), New Orleans, available from http://www.rosemount.com/support/papers/
diagnost0810.doc.
10 SZANYI, R., RATERMAN, M. and ERYUREK, E. Diagnostic capabilities of FOUNDATION fieldbus pressure
transmitters. Hydrocarbon Processing, pp. 53 59, April 2003.
11 HARROLD, D. Protect process instruments to ensure performance. Control Engineering, July 1, 1999.
12 Venturi meters in multiphase flow. Report 286/99 TUV NEL Ltd, East Kilbride, Glasgow.